brand-monitor

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and analyzes user-generated, public content from Reddit via the Reddit JSON API in "Phase 2: Continuous Monitoring" (curl https://www.reddit.com/…/search.json) and also ingests brand websites via web_fetch during onboarding, and those untrusted sources are interpreted by the agent to drive alerts, recommended actions, and automated reply/dig-deeper workflows.